Relative abundance of Mycobacterium in ovine Johne’s disease

No study has determined what proportion of the total microbiota comprises the genus Mycobacterium in ovine Johne’s disease (OJD) tissues. We aimed to assess the relative abundance of Mycobacterium in the ileocaecal lymph node, involved and uninvolved ileal mucosa from sheep with and without OJD, using three extraction methods. Eight sheep, four with and four without OJD, were recruited. Pyrosequencing of the 16Sr RNA gene amplicons for all samples revealed that Mycobacterium represented between 0-92% (average 38%) of the total microbiota of samples from sheep with OJD, and 0-85% (average 13%) of sheep without OJD. Only sheep with OJD had samples that were positive for the IS900 (MAP) element. Mycobacterial strains other than MAP may provide competitive exclusion of MAP and should be further investigated.

Mycobacterium avium subspecies paratuberculosis (MAP) is the etiologic agent of ovine Johne’s disease (OJD), a chronic, contagious granulomatous enteritis of ruminants. The disease causes significant morbidity and mortality worldwide, resulting in significant economic losses¹.

To our knowledge, no study has determined what fraction of the microbial community of non-OJD and OJD affected tissue comprises Mycobacterium. Cheung et al. (2013)² found that Mycobacterium tuberculosis represented less than 1% of the total microbial community of sputum samples of infected patients. Does Mycobacterium comprise such a small fraction of the microbial community in all infectious states and specimens? MAP is a putative trigger of Crohn’s disease in humans, and is detected in some studies but not others, including a recent study of ours³. Is this because it is not being detected?

While other studies have quantified Mycobacterium or MAP in clinical and subclinical specimens^4,5, no study has determined what fraction of the microbial community comprises Mycobacterium in mucosa from sheep with and without OJD. The primary aim is to assess the relative abundance of Mycobacterium in the microbial communities of macroscopically normal bowel mucosa obtained >50 cm downstream the ileocaecal valve (sheep without OJD) or involved mucosa (sheep with OJD), involved (or normal for sheep without OJD) ileal mucosa adjacent the ileocaecal lymph node, and the ileocaecal lymph node of sheep with and without OJD.

Samples of involved and uninvolved ileal mucosa, and ileocaecal lymph nodes were collected from eight sheep, four with and four without OJD, from farms around Rutherglen, Australia. Sterile gloves and implements were used to extract the ileocaecal node before the bowel was opened, so that it did not come into contact with the skin or bowel microbiota. Total DNA was extracted from each sample using three different extraction methods. The first protocol followed the Qiagen DNeasy kit protocol and included an enzymatic lysis step and bead-beating step. The second and third extractions were the same as the first, except the second protocol included a freeze-thaw step, and the third protocol a boiling step. Technical replicates were included for the first protocol. All samples were amplified using barcoded primers targeting the universal bacterial16S rRNA gene, and pyrosequenced using a 454 Genome Sequencer FLX-Titanium platform, as previously described³. Sequences were analysed using Mothur⁶. IS900 PCR was used to determine whether or not Mycobacterium detected in a given sample was MAP, or not.

Four sheep (1, 2, 4 and 5; for example, see Figure 1) were diagnosed with OJD by qualified veterinarians and pathologists, using methods previously described^7,8. Briefly, sheep were initially assessed as having OJD, or not, based on physical signs, including condition, scouring, ill-thrift, and lethargy during mobility. Diagnosis was confirmed by a NATA-accredited pathology laboratory (AgriBio, Victoria), and included MAP culture and DNA analysis, as well as microscopic visualization of acid-fast bacteria using ZN staining . Sheep 1 was initially suspected to have OJD, however negative culture and staining results excluded it from a diagnosis of OJD, and it was found to have metastatic cancer.

Figure 1. Ovine Johne’s disease affected sheep #2, showing physical signs of wasting.

Pyrosequencing showed that Mycobacterium represented between 0–92% (average 38%) of the total microbiota of samples from sheep with OJD, and 0-85% (average 13%) of sheep without OJD (Figure 2). Mucosal samples from sheep 7 were excluded due to the inability to amplify products from them. Sheep 4 and 5 each had a positive lymph node sample for one extraction, comprising 6% and 68% Mycobacterium reads, respectively. These levels were similar to that observed in the mucosal samples from the same animal.

Figure 2. Proportion of involved and normal mucosa and lymph node tissues (sheep 4 and 5) containing Mycobacterium. Sheep with Johne’s disease are coloured orange, without blue. Technical replicates (normal mucosa only) have diagonal texturing. The first numeral in the label indicates the animal number; NM, normal mucosa; IM, involved mucosa; LN, lymph node; the second numeral indicates the DNA extraction: ‘1’, ‘2’ and ‘3’ are replicates, ‘12’ is the technical replicate for normal mucosa (DNA extraction 1) only. Sheep 1, 2, 4 and 5 had OJD, sheep 3, 6 and 8 did not.

The lymph node and mucosa samples of sheep 1 were negative when tested with the MAP-specific primers. Sheep 1 may have had a paucibacillary form of OJD. Samples from all other sheep with OJD were positive, except the normal mucosa sample of sheep 4 (Table 1). All samples from sheep without OJD were negative for the IS900 element.

Table 1. Sheep characteristics, pathology and MAP-specific PCR results

No significant effect of extraction method was detected for either Simpson or Shannon diversity estimates (Simpson Diversity, AOV: P > F = 0.384; Shannon Diversity, AOV: P > F = 0.443), nor technical replicates (Simpson Diversity, AOV: P > F = 0.398; Shannon Diversity, AOV: P > F = 0.566).

Each sample was covered by an average of 2, 644 quality sequences. Six phyla (Firmicutes, Bacteroidetes, Actinobacteria, Proteobacteria, Synergistetes, and TM7) represented ~98% of all sequences (Figure 3). Actinobacteria, of which Mycobacterium is a member, was overrepresented in sheep with OJD, except sheep 1. Microbial community diversity declined as Mycobacterium increased (R²=0.705, p=0.001) (Figure 4).

Figure 3. Proportion of the six dominant bacterial phyla in tissues from sheep with and without ovine Johne’s disease (OJD), showing a predominance of Actinobacteria, of which Mycobacterium is a member, in sheep with OJD. The first numeral in the label indicates the animal number; NM, normal mucosa; IM, involved mucosa; LN, lymph node; the second numeral indicates the DNA extraction: ‘1’, ‘2’ and ‘3’ are replicates, ‘12’ is the technical replicate for normal mucosa only. Sheep 1, 2, 4 and 5 had OJD, sheep 3, 6 and 8 did not.

Figure 4. Figure showing a negative correlation between the Shannon diversity index and the proportion of Mycobacterium in the mucosal microbiota of sheep. Sheep 1, 2, 4 and 5 were diagnosed with OJD.

To our knowledge, this is the first study to describe the full microbial community of mucosa from sheep with and without OJD. Although studies have enumerated MAP in the feces of sheep using RT-PCR^5,9, no study has determined what fraction of the microbial community Mycobacterium comprises, and whether or not these proportions differ in sheep with and without OJD, as well as various tissue types.

Bacterial translocation to the ileocaecal lymph node was observed in 50% (2/4) of sheep diagnosed with OJD. The microbial community of these nodes was similar to that of the involved and normal mucosa from the same sheep. Bacteria in the node are likely to be secondary invaders, present in the node due to a breakdown in the mucosal barrier. Using the same methods described here, we found that bacterial translocation in Crohn’s disease of humans is also non-specific. We also did not detect Mycobacterium in this study using 16S rRNA gene sequencing³. This suggests that Mycobacterium is unlikely to be a cause of Crohn’s disease.

Samples from sheep without OJD were not positive for MAP culture or the IS900 element, but other strains of Mycobacterium were present, representing on average 13% of the total microbiota. Strains of Mycobacterium other than MAP could occupy a niche in the sheep gastrointestinal tract that MAP would otherwise occupy, thereby protecting the animal from colonization of MAP. Future studies should assess the diversity of Mycobacterium strains associated with the mucosa of sheep with and without OJD.

We found the abundance of Mycobacterium, relative to other bacterial taxa in the mucosal samples, was significantly higher in sheep with OJD. Increases in Mycobacterium result in large decreases in Firmicutes. Members of the Firmicutes are important producers of short-chain fatty acids, which are shown to improve intestinal barrier function and repress inflammation¹⁰. Dysbiosis and reduced diversity would likely exacerbate symptoms of OJD, as beneficial bacteria are decreased and their roles in homeostasis unfulfilled.

Mycobacterium makes up a large fraction of the microbial communities of mucosa from sheep with OJD. Future studies should determine if MAP is the sole contributor to these levels of Mycobacterium in sheep with OJD, and assess the diversity of Mycobacterium in sheep with and without OJD. Other Mycobacterium strains may provide competitive exclusion of MAP.